Scott,

On 01/09/2017 07:41 PM, Scott Stobbe wrote:

There where precursor work to Allans Feb 1966 article, but essentially
that where he amalgamed several properties into one to rule them all
(almost). It is indeed the non-convergent properties which motivates a
stronger method. Standard statistics is relevant for many of the basic
blocks, bit things work differently with the non-convergent noise.
Another aspect which was important then was the fact that it was a
counter-based measure. Some of the assumptions is due to the fact that
they used counters. I asked David some questions about why the integral
looks the way it does, and well, it reflects the hardware at the time.

What drives Allan vs. standard deviation is that extra derive function
before squaring
The bias functions that Allan derives for M-sample is really the
behavior of the s-deviation. See Allan variance wikipedia article as
there is good references there for the bias function. That bias function
is really illustrating the lack of convergence for M-sample standard
deviation. The Allan is really a power-average over the 2-sample
standard deviation.

Cheers,
Magnus

Hi

A number of outfits were measuring and spec’ing short term stability in the 1950’s and early 1960’s. Some were doing measures that are pretty close to ADEV. Others were doing straight standard deviation of frequency measurements. Since both got tossed up as “short term stability” confusion was the main result. NIST came in (as it rightly should) and gave us a measurement that does converge. They also spend the next two decades
thumping on a bunch of hard heads to get everybody to use the measurement rather than something with more issues. Once that effort was underway, we got a whole raft of alternatives that each have benefits in certain areas.
ADEV is far from the only measure that could be properly be used today to characterize short term stability.

Bob

Hi Bob,

On 01/09/2017 11:00 PM, Bob kb8tq wrote:

Rather, it provides a basic form of measure for which there exist a
number of estimators. The repairment of the 15 years of troubles was
found inn the MDEV as introduced in 1981, providing a new line of
measure. Improvement in estimator degrees of freedom have rendered
variants. The PDEV provides yet improvement over MDEV.

Cheers,
Magnus

In the late 1960s, Hewlett-Packard engineers worked up a program to have
the 5360A "Computing Pig" (so-called from its weight, 55 pounds without
plug-ins) compute a "fractional frequency standard deviation." It appears
to be similar to the Allen Deviation; I've never figured out the difference
and would appreciate hearing from someone with stronger math skills who can
explain the two.

Jeremy

On Mon, Jan 9, 2017 at 2:00 PM Bob kb8tq kb8tq@n1k.org wrote:

Sent from Gmail Mobile

Hi

The 5360A did ADEV. It only started being called ADEV after a few years had passed.
The 5360A program and it’s various quirks became the topic of a number of post paper
questions in the early 1970’s. The main focus of most of the questions was on bandwidth
limiting ahead of the counter. That question really didn’t get a proper answer for several
more decades.

Bob

For "ADEV" in the 1950's -- follow the words in yellow [1]:

http://leapsecond.com/pdf/1953-Quartz-Greenwich.pdf

For "ADEV" in the 1930's -- I've seen similar treatment done by precision pendulum clock guys. Basically anyone that works with clocks knows to remove initial time offset and constant rate offset when determining the quality of the clock. What's left is the 2nd difference, which is the unexpected deviation in rate. I remember reading a pendulum clock paper from that era where they take the rms of these deviations and call it some kind figure of merit. It goes by other names now, like two-sample variance, sigma(tau), Allan deviation, etc.

/tvb

[1] The original, without OCR or highlight: http://adsabs.harvard.edu/full/1953MNRAS.113...67S

Hi Jeremy,

See the top of page 11 of http://www.hpl.hp.com/hpjournal/pdfs/IssuePDFs/1971-11.pdf where it says:

"Fractional frequency deviation is the term used to describe the frequency
instabilities of a source in the time domain. It has been shown [4] [5] that a
meaningful quantitative measure of fractional frequency deviation is given
by the Allan variance:"
...
4. D. W. Allan, 'Statistics of Atomic Frequency Stan dards,' IEEE Proceedings, Vol. 54, No. 2, 1966.
5. NBS Technical Note 394, 'Characterization of Fre quency Stability.' 

So that's the same as the definition given, for example, here:
http://www.wriley.com/paper2ht.htm#Allan_Variance

There is one source of confusion: ADEV (Allan deviation) is the square root of AVAR (Allan variance). We almost always calculate and plot AVAR. The other confusion is Allan (it's not Allen).

The 5360A is still a favorite of us old-tyme collectors. Lots of information on the old thread (you're in it):

"HP 5360A History?"
https://www.febo.com/pipermail/time-nuts/2016-April/097227.html

The 5360A source code to the ADEV program is here:
http://leapsecond.com/hpj/v22n4/v22n4p10.jpg

/tvb

----- Original Message -----
From: "Jeremy Nichols" jn6wfo@gmail.com
To: "Discussion of precise time and frequency measurement" time-nuts@febo.com
Sent: Monday, January 09, 2017 2:09 PM
Subject: Re: [time-nuts] σ vs s in ADEV

In the late 1960s, Hewlett-Packard engineers worked up a program to have
the 5360A "Computing Pig" (so-called from its weight, 55 pounds without
plug-ins) compute a "fractional frequency standard deviation." It appears
to be similar to the Allen Deviation; I've never figured out the difference
and would appreciate hearing from someone with stronger math skills who can
explain the two.

Jeremy

I mean we almost always calculate and plot ADEV.

Hi Bob,

On 01/10/2017 12:20 AM, Bob kb8tq wrote:

I've not found much on that topic as I've searched. Care to point to a
few papers?

I've been looking at it, and you get somewhat different formulas if you
consider the filter.

Cheers,
Magnus

Hi

It never came up in a paper. It was a question asked from the audience every time the NIST
guys presented an ADEV paper. After a while it got very predictable in terms of who would
stand up and ask what.

Bob